32
1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

Embed Size (px)

Citation preview

Page 1: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

1

Probabilistic Formulation for Skin Detection

Sanun Srisuk42973003

Seminar I

Page 2: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

2

Outline

Problem Statement Literature Review Proposed Skin Detection Experimental Results Conclusions

Page 3: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

3

Problem Statement

Face Detection Methods

•Shape Analysis

•Fuzzy Pettern Matching

•Neural Networks

•SVM

•Hausdorff Distance

Page 4: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

4

Literature Review

Skin Detection using Fuzzy Theory Skin Detection using Color Statistics

Page 5: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

5

Skin Detection using Fuzzy Theory

• Wu et al. [12] propose a method for skin detection using fuzzy theory.

• SCDM and HCDM are skin and hair color models.

• The perceptually uniform color system (UCS) is used for color representation.

Page 6: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

6

SCDM1. Manually select skin regions in each image.2. Prepare a table of 92x140 entries to record the

two dimensional chromatic histogram of skin regions, and initialize all the entires with zero.

3. Convert the chromaticity value of each pixel in the skin regions to UCS, and then increase the entry of the chromatic histogram corresponding to it by one.

4. Normalize the table by dividing all entries with the greatest entry in the table.

Skin Detection using Fuzzy Theory (cont.)

Page 7: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

7

Skin Detection using Fuzzy Theory (cont.)

HCDM

Skin and Hair Color Detectors

Page 8: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

8

Skin Detection using Fuzzy Theory (cont.)

The method proposed in the Wu et al. scheme sometimes fails to detect the real face. Reasons under concern include the followings.•Illumination: This is because, the luminance information is used to detect the hair part of faces, the variance of the illumination color will affect the detection result.•Hairstyle: Faces with special hairstyles, such as skinhead, or wearing a hat, may fail to be detected. This is because the shape of the skin-hair pattern of such a face in the image may become quite different from the head-shape model.

Page 9: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

9

Skin Detection using Color Statistics

Wang et al. [11] present a fast algorithm that automatically detects face regions in MPEG compressed video.•Bayesian minimum rule is used to classify skin or nonskin class.•Classification is performed in YCbCr color model.

Page 10: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

10

Bayesian decision rule

Minimum cost decision rule

Skin Detection using Color Statistics (cont.)

Page 11: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

11

Skin Detection using Color Statistics (cont.)

where

Page 12: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

12

Skin Detection using Color Statistics (cont.)This algorithm can detects 84 of 91 faces (92%), including faces of different sizes, frontal and side-view faces. Detected face regions are marked by white rectangular frames overlaid on the original video frames. There are eight false alarms in this experiment. The algorithm is restricted in several aspects.•It can only be applied to color images and videos, because of the use of chrominance information.•The smallest faces that are detectable by this algorithm are about 48x48 pixels (3x3 macroblocks)

Page 13: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

13

Proposed Skin Detection

In this paper, we•propose a method for color model selection using bayesian estimation.•present an algorithm for color model combination using fuzzy concept.•create 1-D and 2-D histograms for skin pixel classification.

Page 14: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

14

Proposed Skin Detection

probability of skin or nonskin given C1 and C2

where

denotes the skin and nonskin classes.

be the 2-D histogram of skin and nonskin areas.

are chrominance components.

Page 15: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

15

Proposed Skin Detection

represents the probability of given class

is the a priori probability of class

Page 16: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

16

Proposed Skin Detection (cont.)

Maximum a posteriori (MAP)

A decision function for selecting the chrominance

be the selected chrominance components.

Page 17: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

17

Proposed Skin Detection (cont.)

Page 18: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

18

Proposed Skin Detection (cont.)

membership function

Page 19: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

19

Proposed Skin Detection (cont.)

Skin detection function

where

is the skin color likeness function.

are the minimum and maximum values of the range of human skin in selected chrominance components.

is the range from a to b and from c to d.

Page 20: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

20

Proposed Skin Detection (cont.)

is the weighting coefficient associated with chrominance component

is the probability generated by 1-D histogram.

Page 21: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

21

Results

Skin detection under varying illuminations

Page 22: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

22

Results

Skin detection under different races

Page 23: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

23

Results

Original Image Our proposed method

Page 24: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

24

Results

HSV [6] YCbCr [3]

Page 25: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

25

Results

Original Image Our proposed method

Page 26: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

26

Results

HSV [6] YCbCr [3]

Page 27: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

27

Results

Original Image

Our proposed method

Page 28: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

28

Results

HSV [6]

YCbCr [3]

Page 29: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

29

Results

Original Image

Our proposed method

Page 30: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

30

Results

HSV [6]

YCbCr [3]

Page 31: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

31

Conclusions

The skin and nonskin probabilities are created from 1-D and 2-D histograms.

Bayesian estimation is used to select appropriate well-known color models.

Skin detection is performed by fuzzy membership function and normalized by 1-D histogram.

The method is proposed for robust skin detection under varying illuminations and different races.

Page 32: 1 Probabilistic Formulation for Skin Detection Sanun Srisuk 42973003 Seminar I

32